religious identity and economic behavior 2010.04.13

31
Religious Identity and Economic Behavior * Daniel J. Benjamin Cornell University, Institute for Social Research, and NBER James J. Choi Yale University and NBER Geoffrey Fisher Cornell University Current draft: May 1, 2010 Abstract Although many scholars (e.g., Weber, 1930) have hypothesized that religious identity norms affect economic outcomes, empirical tests have been hampered by the difficulty of identifying exogenous variation in religion. We create exogenous variation by randomly varying religious identity salience in laboratory subjects. The marginal effect of religious identity is the change in subjects’ choices when religion is salient. We test six hypotheses from prior literature. We find that Protestantism increases contributions to public goods. Catholicism decreases contributions to public goods, decreases expectations of others’ contributions to public goods, and decreases risk aversion. Judaism increases worker reciprocity in a bilateral labor market gift-exchange game. We find no evidence of religious identity effects on disutility of work effort, discount rates, or generosity in a dictator game. JEL Classification: C91, C92, Z12 Keywords: religion, identity, norm, gift exchange, public goods, dictator game * We thank Azim Shariff and Ara Norenzayan for sharing with us the religious identity priming instrument. We are grateful to Stefano DellaVigna, Kirabo Jackson, Ted O’Donoghue, and participants at the NBER’s Economics of Religion Conference, UC Santa Barbara’s Behavioral/Experimental Economics Conference, and the University of Maryland’s Labor/Public Economics Seminar for helpful comments and suggestions. We thank Mario Basora, Maria Bodiu, Kristin Brinley Conner, Evan Buntrock, Jim Casteleiro, John Farragut, Isabel Fay, Joshua Funt, Arjun Gokhale, Jesse Gould, Rebecca Hausner, Ben Hebert, Liying Huang, Ahmed Jaber, Bige Kahraman, Anqi Kang, Philip Kauders, June Kim, Xiaoying Lin, Michael Luo, Max Mihm, Gregory Muenzen, Christopher Nieves, Collin Raymond, Alex Rees-Jones, John Schemitsch, Nathaniel Schorr, Dennis Shiraev, Nichole Szembrot, Russell Toth, Elizabeth Truax, and Ryan Yamada for their research assistance. We thank the National Institute on Aging (grants P30-AG012810 and P01-AG005842) for financial support. E-mail: [email protected], [email protected], [email protected].

Upload: others

Post on 12-Sep-2021

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Religious identity and economic behavior 2010.04.13

Religious Identity and Economic Behavior*

Daniel J. Benjamin Cornell University, Institute for Social Research, and NBER

James J. Choi

Yale University and NBER

Geoffrey Fisher Cornell University

Current draft: May 1, 2010

Abstract Although many scholars (e.g., Weber, 1930) have hypothesized that religious identity norms affect economic outcomes, empirical tests have been hampered by the difficulty of identifying exogenous variation in religion. We create exogenous variation by randomly varying religious identity salience in laboratory subjects. The marginal effect of religious identity is the change in subjects’ choices when religion is salient. We test six hypotheses from prior literature. We find that Protestantism increases contributions to public goods. Catholicism decreases contributions to public goods, decreases expectations of others’ contributions to public goods, and decreases risk aversion. Judaism increases worker reciprocity in a bilateral labor market gift-exchange game. We find no evidence of religious identity effects on disutility of work effort, discount rates, or generosity in a dictator game. JEL Classification: C91, C92, Z12 Keywords: religion, identity, norm, gift exchange, public goods, dictator game * We thank Azim Shariff and Ara Norenzayan for sharing with us the religious identity priming instrument. We are grateful to Stefano DellaVigna, Kirabo Jackson, Ted O’Donoghue, and participants at the NBER’s Economics of Religion Conference, UC Santa Barbara’s Behavioral/Experimental Economics Conference, and the University of Maryland’s Labor/Public Economics Seminar for helpful comments and suggestions. We thank Mario Basora, Maria Bodiu, Kristin Brinley Conner, Evan Buntrock, Jim Casteleiro, John Farragut, Isabel Fay, Joshua Funt, Arjun Gokhale, Jesse Gould, Rebecca Hausner, Ben Hebert, Liying Huang, Ahmed Jaber, Bige Kahraman, Anqi Kang, Philip Kauders, June Kim, Xiaoying Lin, Michael Luo, Max Mihm, Gregory Muenzen, Christopher Nieves, Collin Raymond, Alex Rees-Jones, John Schemitsch, Nathaniel Schorr, Dennis Shiraev, Nichole Szembrot, Russell Toth, Elizabeth Truax, and Ryan Yamada for their research assistance. We thank the National Institute on Aging (grants P30-AG012810 and P01-AG005842) for financial support. E-mail: [email protected], [email protected], [email protected].

Page 2: Religious identity and economic behavior 2010.04.13

2

At least since Weber (1930), scholars have hypothesized that norms tied to religious

identities affect economic outcomes. Weber argued that Protestantism encouraged capital

accumulation and a strong work ethic, thus leading to the rise of capitalism. More recently, Barro

and McCleary (2003, 2006) find evidence that belief in heaven and hell increases GDP growth

rates, a result that they hypothesize is due to the salutary effect of this belief on work ethic,

honesty, trust, and thrift. Putnam (1993) and La Porta et al. (1997) argue that Catholicism

inhibits trust, which has negative effects on GDP growth, government efficiency, the production

of public goods, and the maximum feasible size of corporations. Putnam (1993) also argues that

Protestantism promotes trust. Relatedly, Ruffle and Sosis (2007) find that participation in

collective religious rituals is associated with greater trust and cooperativeness. Stulz and

Williamson (2003) show that a country’s principal religion is correlated with the strength of its

creditor rights; Guiso, Sapienza, and Zingales (2003) find positive correlations between Christian

religions and attitudes conducive to economic growth; and Hilary and Hui (2009) and Kumar,

Page, and Spalt (2009) argue that religious risk norms affect corporations’ investment decisions

and individuals’ stock portfolios.1

However, causal inference about the effect of religious identity norms has been hampered

by the difficulty of identifying exogenous variation in religious identities. Religious beliefs are

endogenous choices and are likely to be correlated with many unobserved factors that have

causal impacts on economic outcomes.

In this paper, we create exogenous variation in the strength of religious identity norms.

According to “self-categorization theory,” a long-standing idea in psychology (James, 1890;

Turner, 1985), each person belongs to multiple social categories, such as religion, gender, and

occupation, which each has its own set of norms. Behavior in a given moment is more

powerfully affected by the norms of categories that are salient than the norms of categories that

are not salient. If an environmental cue, or a “prime,” makes a certain category temporarily more

salient, behavior shifts towards the salient category’s norm. We can therefore identify the

marginal behavioral effect of religious identity norms by making religious identity salient to a 1 There is also a large literature finding religiosity has a positive correlation with salutary individual outcomes, which could be due to the causal impact of religious norms. These outcomes include higher educational attainment, higher income, lower levels of welfare receipt and disability, higher marriage probability, lower divorce probability, better health, greater self-reported happiness, and greater resilience to childhood disadvantage (e.g. Freeman, 1986; Ellison, 1991; Levin, 1994; Gruber, 2005; Dehejia et al., 2009). Becker and Woesmann (2009) argue that most of the Catholic-Protestant prosperity gap in late-19th-century Prussia can be accounted for by higher literacy among Protestants, driven by Protestants’ desire to read the Bible. Iannaccone (1998) surveys much of this literature.

Page 3: Religious identity and economic behavior 2010.04.13

3

randomly selected subset of laboratory subjects and observing how these primed subjects’

subsequent behavior differs from the behavior of unprimed subjects. This methodology has

previously been used to identify economic effects of racial, ethnic, and gender identity norms by

Benjamin, Choi, and Strickland (forthcoming).

Using priming techniques, we examine six hypotheses derived from prior literature:

H1. Contributions to public goods are affected positively by Protestantism and negatively by

Catholicism (Putnam, 1993; La Porta et al., 1997).

H2. Trust is affected positively by Protestantism and negatively by Catholicism, and trust is

the mechanism explaining the relationship between religion and public goods

contributions (Putnam, 1993; La Porta et al., 1997).

H3. Financial risk-taking is reduced by religion. Kumar, Page, and Spalt (2009) argue that

risk-taking is reduced by Protestantism, whereas Hilary and Hui (2009) argue that risk-

taking is reduced by both Protestantism and Catholicism.

H4. Thrift and capital accumulation are promoted by religion (Weber, 1930; Guiso, Sapienza,

and Zingales, 2003; Barro and McCleary, 2003, 2006).

H5. Generosity is increased by religion (Friedrichs, 1960; Batson, Schoenrade, and Ventis,

1993).

H6. Work ethic is increased by religion, and especially by Protestantism (Weber, 1930; Barro

and McCleary, 2003, 2006).

To operationalize these hypotheses in a laboratory setting, we map them into seven outcomes

that can be measured using standard experimental methods. The first five outcomes are

contributions to a laboratory public good, expectations about others’ contributions to the

laboratory public good, elicited risk aversion, elicited discount rates, and the amount given away

in a dictator game. Since work ethic could be interpreted as a low disutility of effort or as a high

reciprocity motive toward an employer when labor contracts are incomplete, we measure two

outcomes corresponding to each of these concepts: the number of anagrams attempted at a

constant piece-rate wage, and the rate at which effort expended as a worker rises with wage

offers in a labor market bilateral gift-exchange game.

Page 4: Religious identity and economic behavior 2010.04.13

4

Consistent with H1, we find that Protestantism increases contributions to the laboratory

public good, while Catholicism decreases contributions. In partial support of H2, we find that

Catholicism decreases expectations of others’ contributions to the public good, and Catholicism

has no statistically significant effect on contributions once its impact on trust is controlled for.

However, Protestantism does not affect trust, suggesting that Protestantism’s positive impact on

public good contributions comes from a norm to unconditionally contribute to public goods.

Our data do not support H3. In accordance with Kumar, Page, and Spalt (2009), we do

find that religious identity norms cause Protestants to become relatively more risk averse than

Catholics, but this is because Catholicism increases risk-taking, rather than Protestantism

reducing risk-taking as Kumar, Page, and Spalt (2009) hypothesize.

We find no evidence for Judeo-Christian identity effects on discount rates and generosity

(H4 and H5). Nor do we find Christian work ethic effects (H6). However, we do find a work

ethic effect for Judaism; among Jews, priming religion increases the rate at which workers

increase their effort in response to higher wages in the gift-exchange game.

We are aware of three prior psychology papers that manipulate religious identity salience

and measure changes in subsequent behavior. Shariff and Norenzayan (2007) find that priming

religion increases generosity in a dictator game. Even though we use the same priming

instrument they do, we are unable to replicate their result in our much larger sample. None of our

dictator game treatment effects are statistically significant, and the point estimates indicate that

dictator game generosity, if anything, slightly decreases when religion is made salient. Toburen

and Meier (forthcoming) use the Shariff and Norenzayan (2007) priming instrument and find that

religiously primed subjects given unsolvable anagrams to solve as part of a “verbal intelligence

test” (for which they were not paid) spend more time on the task. Randolph-Seng and Nielsen

(2007) find that priming religion reduces the frequency of high performance in an unmonitored

laboratory task, which they interpret as a reduction in cheating.2 In the economics literature,

Hilary and Hui (2009) find suggestive experimental evidence (not reported separately by

religion) that priming religion increases risk aversion in hypothetical risk choices. In our larger

sample using incentivized choices, we find evidence in the opposite direction for Catholics.

2 There have been many other studies examining the effect of making non-religious identities salient, including Reicher and Levine (1994), Forehand, Deshpandé, and Reed II (2002), and LeBoeuf, Shafir, and Bayuk (2010).

Page 5: Religious identity and economic behavior 2010.04.13

5

There are also a number of papers that report correlations between religious affiliation and

behavior in economics experiments (e.g., Anderson and Mellor, 2009; Tan and Vogel, 2008).

Our paper proceeds as follows. Section I describes a theoretical framework for

understanding how priming effects allow us to make inferences about norms. Section II describes

the pilot experiment we used to confirm that our priming instrument has the desired effect on

identity salience. Section III describes the methodology of our main experiment. Section IV

presents the main experiment’s empirical results, including a joint hypothesis test to address

concerns about Type I error due to multiple hypothesis testing. Section V concludes.

I. A Theoretical Framework

Within our theoretical framework, which is inspired by Akerlof and Kranton (2000) and

developed in Benjamin, Choi, and Strickland (forthcoming), priming a social category reveals

the marginal effect of increasing the strength of affiliation with that category. Let x be some

choice, such as how much to contribute to a public good or how much to trust a stranger. An

individual belongs to a social category C, such as Protestantism, with strength s ≥ 0. Let x0

denote the individual’s preferred choice in the absence of identity considerations, and let xC

denote the choice prescribed for members of social category C. The individual chooses x to

maximize

U = –(1 – w(s))(x – x0)2 – w(s)(x – xC)2, (1)

where 0 ≤ w(s) ≤ 1 is the weight placed on the norm for social category C in the person’s

decision. We assume that w(0) = 0 and w′ > 0. Deviating from the norm prescribed for one’s

category causes disutility that is increasing in s. We assume that s has a steady-state value s but

can be temporarily increased to s + ε by a category prime, where ε > 0.

The first-order condition of (1) gives the optimal action, x*(s) = (1 – w(s))x0 + w(s)xC,

which is a weighted average of the preferred action without identity considerations and the

category norm. This condition yields several implications that guide our analysis.

First, the higher the steady-state strength s of the category affiliation, the closer x* is to

xC in steady state. Second, a category prime also causes x* to move closer to xC. Thus, the

behavioral effect of priming social category C reveals the marginal behavioral effect of

increasing the steady-state strength s of C. This is why priming manipulations are a useful

experimental procedure for studying how identity affects steady-state preferences. Third, the sign

Page 6: Religious identity and economic behavior 2010.04.13

6

of the priming treatment effect, x*(s + ε) – x*(s) ≈ (dx*/ds)ε = w′(s)(xC – x0)ε, depends on the sign

of xC – x0. Even if the s , x0, and w(·) of an experimental sample differ from those of the general

population, the directional effects of priming the sample will generalize to the population as long

as xC – x0 has the same sign for both groups.

Finally, although the direction of the priming effect reliably identifies the sign of xC – x0,

differences in the priming effect’s magnitude across people may arise through a number of

channels. Assume without loss of generality that xC > x0. Priming will have a larger effect if the

identity norm is more extreme (i.e., xC is larger) or the person’s preferred action in the absence of

identity considerations is more extreme in the opposite direction (i.e., x0 is smaller). Priming will

also have a larger effect if the salience manipulation is more effective at increasing identity

salience for that particular person (i.e., ε is larger) or the person’s choices are more sensitive to a

given change in identity salience (i.e., w′ is larger). This latter difference can arise either because

the w function has a different shape or because the person has a different steady-state s , so that

the points at which he or she evaluates the w function differ. Priming will have no effect if there

is no identity norm for choices in the measured domain, in which case the person always chooses

x0 regardless of identity salience.

II. Validating the Priming Instrument

The priming instrument, first used by Shariff and Norenzyan (2007) to study the effect of

priming religious concepts, is a sentence-unscrambling task where subjects are asked to drop the

irrelevant word in a five-word group and rearrange the remainder to form a four-word sentence.

For example, “yesterday it finished track he” becomes “he finished it yesterday.” Each subject

unscrambles ten sentences.

The sentences vary depending on whether the subject is in the religion-salient condition

or the control condition. Five of the sentences unscrambled by religion-salient subjects contain

religious content. These five sentences are: “she felt the spirit,” “the dessert was divine,” “give

thanks to God,” “the book was sacred,” and “prophets reveal the future.” None of the control

subjects’ sentences contain religious content. An advantage of this priming instrument is that it is

subtle; compared with blatant primes, subtle primes more reliably cause behavior to conform to

norms (Wheeler and Petty, 2001).

Page 7: Religious identity and economic behavior 2010.04.13

7

We recruited 91 students at the University of Michigan for a pilot experiment to confirm

that the priming instrument increases the strength of religious identity affiliation. Subjects were

randomly assigned to complete the religion-salient task or the control task. Subjects were not

aware that this task differed across subjects. Immediately after the sentence unscramble, the

questionnaire asked: “What five aspects of your identity (such as ‘male/female’ or ‘college

student’) are most important to you?” Forty-seven percent of subjects in the religious-salient

condition listed a religious identity in response, compared to only 25 percent of subjects in the

control condition. This difference is significant at the 3 percent level (z-statistic = 2.16).

In contrast, for each of the other identities listed by subjects (in descending order of

frequency, they were: gender, student, nationality / ethnicity, personality trait, family

relationship, professional interest, non-family relationship, age group, activity group, political

orientation, physical feature, and socioeconomic class), the p-value for a test of differences

across conditions in the proportion of subjects who mentioned it varied from 25 to 97 percent.

This evidence suggests that the priming instrument increases religious identity salience without

affecting the salience of other identities.

III. Main Experiment Procedure

Participants in the main experiment were 827 Cornell University students. To avoid

making religious identity salient to all the subjects, we did not mention in our recruiting

materials that we were looking for people of particular religions. Sessions were administered by

computer, most using the program z-Tree (Fischbacher, 2007) but some using VBA for

Microsoft Excel, depending on which dependent variables we measured in the session.

We randomly assigned subjects to complete the religion-salient or control sentence

unscramble. Subjects were not aware that this task differed across subjects. After completing the

sentence unscramble, they participated in strategic games, incentive-compatible preference

elicitations, and an anagram-solving task. We describe the games, elicitations, and anagram task

below. Subjects were told at the beginning of the experimental session that any interactions they

had with other subjects would be anonymous one-shot interactions. In order to avoid excessively

long sessions, each subject engaged in only a subset of the post-unscramble tasks. The order in

which the post-unscramble tasks appeared varied across sessions.

Page 8: Religious identity and economic behavior 2010.04.13

8

A. Public goods game

We measured the willingness to contribute to a public good by assigning each subject to a

group of four and endowing him or her with $1.3 Subjects could contribute any fraction of their

dollar to a group account, which is the laboratory public good. Contributions were doubled and

then distributed evenly among the four group members. Subjects kept any money that they did

not contribute. Total group earnings are maximized (at $2 per group member) if each member

contributes his or her entire dollar to the group account. However, in the absence of other-

regarding preferences, it is a dominant strategy to contribute nothing, since the private return on

a contribution is –50 percent.

Before eliciting subjects’ own contributions, we asked subjects to give their best guess of

how much the other three members of their group would contribute on average—a measure of

trust.4 Subjects’ payments did not depend upon the value of this guess.

Existing evidence suggests that behavior in laboratory public goods games like ours is

associated with contributions to public goods outside the laboratory. Carpenter and Seki

(forthcoming) find that cooperativeness in a laboratory public good game is positively correlated

with the productivity of Japanese fishermen who pool their catch with other fishermen, and Fehr

and Leibbrandt (2008) find that Brazilian fishermen who contribute more to a laboratory public

good are less likely to over-exploit their common fishing ground. Laury and Taylor (2008) find

that the amount contributed to a laboratory public good is positively correlated with the

willingness to contribute money to a local tree-planting organization.5

B. Risk aversion elicitation

We elicited small-stakes risk preferences by asking participants to make six binary

choices between $1 for sure and a 50 percent chance at a larger amount, ranging from $1.60 to

3 Marwell and Ames (1979) were the first to run this type of experiment to study the willingness to contribute to public goods. 4 We use the term “trust” in the sense of La Porta et al. (1997, p. 333): trust is when “people expect certain fair or cooperative behavior of their opponents even when they do not expect to see them again.” See also Gächter, Herrmann, and Thöni (2004) for an explication of how trust applies to a simultaneous-move public goods game. 5 However, Laury and Taylor (2008) find that this relationship is driven by the most selfish subjects, and that differences in contributions among less selfish subjects do not reliably predict differences in contributions to the field public good. This may be because the choices about contributing to the laboratory public good and the field public good occurred close together in time. Virtuous behavior tends to encourage more selfish behavior subsequently due to the “licensing effect” (Kahn and Dhar, 2006); because an individual has bolstered his self-concept as a virtuous person through the initial virtuous act, he is freed to act selfishly afterwards.

Page 9: Religious identity and economic behavior 2010.04.13

9

$3.60. We paid subjects for all six choices in accordance with their stated preferences; for

example, if a subject selected the sure $1 payment if and only if winning the gamble yielded

$3.60, then she would have a 50 percent chance of earning $5 and a 50 percent chance of earning

$5 + $3.60 = $8.60 in this section. We measured larger-stakes risk preferences with six

analogous choices, where the monetary amounts were 100 times larger than the small-stakes risk

choice amounts and there was only a small chance that the subject’s choices would be

implemented for payment.6 The section’s instructions made it clear that the questions were not

intended to evaluate subjects normatively: “It’s important to keep in mind that there are no right

or wrong answers here. Which choice you make is a matter of personal preference.” We used this

same wording again in the instructions for the discount rate elicitations.

Risk aversion measures derived from incentive-compatible experimental choices such as

ours are highly correlated with measures from hypothetical choices, which in turn predict risky

behaviors such as smoking, drinking, failing to hold insurance, holding stocks rather than

Treasury bills, being self-employed, switching jobs, and moving residences.7

C. Discount rate elicitation

We measured time preferences by asking participants to make 12 binary choices between

receiving $10 now and receiving some larger amount one week from now, and another 12 binary

choices between receiving $10 one week from now and receiving some larger amount two weeks

from now. The larger delayed amounts ranged from $10.10 to $15. After a subject made these

choices, we randomly selected one of the 24 choices and paid the subject according to his or her

stated preference in this choice. All payments were made by a check given to the participant

immediately following the experiment. Delayed payments were implemented via post-dated

check.8

Our approach to measuring time preferences is standard (Frederick, Loewenstein, and

O’Donoghue, 2002). Similar measures predict variation in discounting-related behaviors such as

drug addiction, cigarette smoking, excessive gambling, use of commitment savings devices,

6 One of the six large-stakes risk choices was randomly chosen to be paid out if the subject could correctly predict two spins of a roulette wheel, which implied a one in 1,444 chance of one large-stakes risk choice being implemented. 7 See Barsky et al. (1997), Guiso and Paiella (2008), Dohmen et al. (2005), and Sahm (2007). 8 If the subject received a delayed payment for this section, then earnings from other sections were paid through a separate check that was immediately cashable.

Page 10: Religious identity and economic behavior 2010.04.13

10

borrowing on installment accounts and credit cards, rapid exhaustion of food stamps, delayed

application to an MBA program, and defaulting on loans.9

D. Dictator game

In our implementation of the dictator game (Kahneman et al., 1986; Forsythe et al.,

1994), we endowed each subject with $1 and randomly assigned him or her to another participant

in the session. The subject could choose to give any portion of that $1 to the other subject. A

profit-maximizing individual would keep the entire dollar for himself, so the amount given away

is a measure of pure altruism. Benz and Meier (2008) find that generosity in laboratory dictator

games is positively correlated with charitable giving outside the laboratory.

E. Labor market tasks

Work ethic can be interpreted in terms of economic theory as relating to an individual’s

disutility of effort, which determines the willingness to exert a contracted-upon amount of effort

at a given wage rate, or the strength of the reciprocity motive toward an employer that causes a

worker to supply more effort in response to a higher wage when the labor contract is incomplete.

We measure identity effects on both variables.

To measure identity effects on the disutility of effort, we asked subjects to solve as many

four-letter anagrams as they could in a five-minute period. We paid participants five cents per

correctly solved anagram. Because the piece-rate wage we offered is a contingent payment,

subjects’ reciprocity norm should not have been strongly activated in this task. Hence, the

quantity of effort supplied by subjects equates the marginal cost of effort with the marginal

benefit of the expected payment from anagram-solving effort. Shifts in the amount of effort

exerted across salience conditions reflect shifts in the marginal effort cost function.10 By

9 See Fuchs (1982), Bickel, Odum, and Madden (1999), Kirby, Petry, and Bickel (1999), Petry and Casarella (1999), Kirby and Petry (2004), Shapiro (2005), Ashraf, Karlan, and Yin (2006), Meier and Sprenger (2010), and Reuben, Sapienza, and Zingales (2009). 10 In mathematical notation, let e be the quantity of effort supplied, with units normalized so that the expected number of anagrams solved in the five minutes allotted is equal to e. Let p be the payment in the numeraire good per anagram solved, and let the increasing convex function c(e) denote the utility cost of supplying effort. The subject’s utility function is U(e; p) = ep – c(e). As long as the optimal amount of effort e* is interior, it is characterized by the first-order condition, c′(e*) = p. Since p is the same across the treatment and control groups, changes in effort supply can be attributed to changes in the c′ function.

Page 11: Religious identity and economic behavior 2010.04.13

11

restricting the task length to five minutes, we reduced the chances that primed subjects’ religious

identity salience would fade during the task.11

We measured work provision in an incomplete contractual setting by running a labor

market bilateral gift-exchange game (Fehr et al., 1998), modeled on one of the implementations

of Charness, Frechette, and Kagel (2004). We paired subjects together and told them that one of

them would play the role of the Manager and the other the role of the Employee. We described

the roles in these terms so that subjects would be more likely to use norms applicable to labor

markets in their choices. After playing once, subjects played the game again, but this time in the

opposite role and with a different partner.

In the first stage of the game, the Manager paid a wage to the Employee between $0 and

$4 that is a multiple of 50 cents. The Manager could not change the wage later. In the second

stage, the Employee saw the wage and chose a work quantity to supply that is an integer between

1 and 10, inclusive. Employees were told that the Manager would be shown their work quantity

choice. The Employee’s cost of work provided was an increasing convex function of work

quantity: $0.00, $0.04, $0.08, $0.16, $0.24, $0.32, $0.40, $0.48, $0.60, and $0.72 as work

quantity rose from 1 to 10. The Employee’s earnings were the wage received minus the cost of

work provided. The Manager’s earnings were ($4 – wage paid) × work quantity provided by the

Employee ÷ 10.

To facilitate calculation, we provided subjects a lookup table that showed the Manager

and Employee’s earnings at each wage and work quantity combination. Managers simply chose

one wage, while Employees indicated a contingent work supply choice for each possible wage.

After observing the Managers’ wage choice, we implemented the Employees’ work supply

choice based on that wage. The profit-maximizing strategy for the Employee is to always supply

the minimum amount of work, since the Manager is unable to contract on effort and has no

opportunity to punish the Employee for shirking. If the Manager believes the Employee is a

profit maximizer, his or her own profit-maximizing response is to offer a $0 wage. Despite these

equilibrium predictions, the prior literature finds that, consistent with the presence of a

reciprocity norm, Employees usually supply positive effort that is increasing in the wage offer.

Managers anticipate this norm and rarely offer the minimum possible wage.

11 Subjects given the anagram-solving task performed no other tasks after the sentence unscramble.

Page 12: Religious identity and economic behavior 2010.04.13

12

Because the Employee’s work quantity is the choice of a number rather than effort in a

real work task, and because the cost of providing this work quantity is determined by a function

that is the same for all subjects, differences in work provided across salience conditions are

driven solely by changes in the strength of subjects’ reciprocity norm and not by changes in the

utility cost of effort.

F. Debriefing questionnaire

At the end of the session, after payoffs had been revealed, subjects completed a

debriefing questionnaire that collected information about their demographic characteristics,

beliefs about the experiment, and religious beliefs. We also included numerous decoy questions

to mask the purpose of the study, so that subjects would not contaminate future subjects by

telling them that we were running an experiment about religion. We discuss responses to the

relevant questions in further detail in Section IV.

IV. Main Experiment Results

A. Sample selection

Our sample consists of 264 Protestant/other Christians (whom we refer to collectively as

“Protestants”), 199 Catholics, 95 Jews, and 269 atheists/agnostics.12 In the debriefing

questionnaire, we asked subjects, “What do you think this study is about?” To avoid estimating

treatment effects that are driven by experimenter demand effects, the above sample of 827

excludes four subjects who guessed that the study had something to do with religion.13 In

addition, we have dropped four subjects who left more than half of the sentence-unscramble

responses blank, since they were likely not to have been properly primed. Our results are not

sensitive to including these dropped subjects.

12 Since we analyze results separately by religion, it would be problematic if the priming manipulation affected how participants categorized themselves, e.g., causing marginal atheist/agnostics to identify themselves with the religious tradition in which they were raised. We find that religious affiliations are almost perfectly balanced across treatment and control groups. In the treatment group, there are 130 Protestants, 101 Catholics, 43 Jews, and 136 atheists/agnostics, compared with 134, 98, 52, and 133 in the control group, respectively. 13 At the end of the experiment, we also asked subjects whether they believed that their experimental choices would affect their payments as the instructions specified. Ninety percent of subjects reported believing these payment promises. Among the subjects who participated in the dictator, public goods, and gift exchange games, 88 percent reported believing that their choices would affect other participants’ payments exactly as we had specified.

Page 13: Religious identity and economic behavior 2010.04.13

13

B. Public goods game

Panel A of Table 1 shows coefficients from regressing the amount contributed to the

public good on a dummy for being in the religion-salient condition. The constant coefficients

indicate that among unprimed subjects, Catholics contribute the most on average, agnostics and

atheists contribute the least, and Protestants and Jews are in the middle. However, selection into

our sample of Cornell students is not random. And even if our sample were representative of

each religion’s members, the many unobserved variables that are correlated with religious

affiliation would prevent us from inferring the causal effect of religion by simply comparing

subject choices across religions. To learn about the impact of religion, we instead turn to

comparisons between the treatment and control groups within each religion.

We find, consistent with the conclusions of Putnam (1993) and La Porta et al. (1997), that

Protestantism increases the supply of public goods while Catholicism suppresses it. Protestants

for whom religious identity is salient contribute 15 cents more to the public good than control

Protestants, whereas primed Catholics decrease their contributions by 18 cents. Jewish subjects’

contributions are unaffected by the prime. Interestingly, the prime has a marginally significant

effect on atheists and agnostics, increasing their contributions by 12 cents (p-value = 0.08). This

could be due to the existence of an atheist or agnostic norm regarding public good contributions,

but it may alternatively arise from the activation of residual religious norms present among

subjects who were raised in a faith and have subsequently abandoned it.

Putnam (1993), La Porta et al. (1997), and Ruffle and Sosis (2007) argue that the channel

through which religion affects public goods provision is trust. Among Catholics, our data support

this hypothesis. Panel B of Table 1, which contains regressions of subjects’ expectations of other

group members’ average contribution on a religion-salient dummy, indicates that relative to

control Catholics, primed Catholics expect the average member of their group to contribute 12

cents less. Primed atheists and agnostics expect others to contribute 9 cents more, a marginally

significant result. However, there is no significant effect of priming on Protestant expectations.

This suggests that Protestantism’s positive effect on public good contributions in Panel A is due

to a Protestant norm to unconditionally contribute to public goods, rather than a Protestant norm

to trust others.

Page 14: Religious identity and economic behavior 2010.04.13

14

Panel C shows that subjects’ reported expectations are not just cheap talk. The

coefficients in this panel are from a regression of contributions on a religion-salient dummy and

the subject’s expectation of other group members’ average contribution. We find that subjects’

own contributions increase almost one-for-one with their expectations.14 Once expectations are

controlled for, primed Catholics do not contribute significantly less than unprimed Catholics,

indicating that the Catholic priming effect on contributions operates entirely through

Catholicism’s negative effect on trust. Similarly, the positive priming effect on atheist and

agnostic contributions is no longer significant after controlling for expectations. In contrast,

primed Protestants contribute a highly significant 12 cents more even after controlling for

expectations. The Protestant priming coefficient is statistically distinguishable from the Catholic

priming coefficient at the 5 percent level.

C. Risk aversion

In the risk preference regression, our dependent variable is the minimum risk premium—

that is, the expected return offered by the gamble in excess of the risk-free return—that the

subject requires to accept the gamble. For example, if the subject would choose to gamble for a

50 percent chance of receiving X rather than accept a sure $1 if X is at least $2.40, then the

reservation risk premium is (2.40 × 0.5 – 1)/1 = 0.20.15 Because we observe binary choices over

only a finite number of risk premia, we use an interval regression (Stewart, 1983), which is a

generalization of the tobit estimator that accommodates dependent variable values that are not

precisely observed but are known to lie somewhere within an interval. We observe two risk

premia for each subject—one for the small-stakes gamble and one for the larger-stakes gamble—

so each subject appears in the regression twice, and we cluster standard errors by subject (Froot,

1989; Rogers, 1993).

14 In untabulated regressions, we find that this relationship does not differ significantly between primed and unprimed subjects in any of the religious groups. 15 In this formula, we treat the risk choice as investing a $1 endowment into either the risk-free or risky asset. Our decision to use the “minimum” risk premium as the dependent variable instead of the maximum risk premium that induces a safe choice only matters when participants’ risk choices are inconsistent, switching between choosing the safe option and the risky option at more than one value of X. Since 95 percent of participants behaved consistently, our results are virtually identical if we use the “maximum” risk premium instead. Similarly, our choice to use the “minimum” interest rate in our time preference regressions in Section IV.D does not matter because 97 percent of participants had a unique interest rate switch point.

Page 15: Religious identity and economic behavior 2010.04.13

15

Table 2 shows that priming religious identity causes the average risk premium required to

entice agnostics and atheists to forego a sure payout to fall by 12 percentage points. For the

small-stakes gamble, this represents a fall from 16 percent to 4 percent. We also find evidence

that religious identity salience causes Catholics’ risk premium to fall by 11 percentage points.

This is a drop from 21 percent to 10 percent for the small-stakes gamble. The effect is significant

only at the 10 percent level in this regression specification, but if we additionally control for a

gender dummy (which is highly significant), the priming effect point estimate becomes

significant at the 5 percent level. We find no significant identity effects on risk-taking for

Protestants and Jews.

Our results are consistent with Kumar, Page, and Spalt’s (2009) conclusion that religious

norms raise the risk-taking of Catholics relative to Protestants. But whereas they argue that this

effect occurs because of risk-averse Protestant norms, we find that the effect is instead driven by

risk-seeking Catholic norms. Our two negative Christian priming point estimates are inconsistent

with Hilary and Hui’s (2009) argument that both Protestantism and Catholicism discourage risk-

taking.16

D. Discount rate

In the time preference regression, our dependent variable is the log of the minimum

continuously compounded weekly interest rate that the subject requires to choose the later

payment. That is, we apply the log operator once to transform the reservation gross interest rate

into the continuously compounded net interest rate, and then we apply the log operator again.

The second application of the log operator causes the estimator we describe below to assume that

reservation continuously compounded interest rates are conditionally log-normal, thus ruling out

negative discount rates. Each subject appears in the regression twice because we have two

discount rate observations: now versus one week in the future, and one week versus two weeks

in the future. Therefore, we cluster our standard errors by subject.

As in the risk preference elicitation, we only observe binary choices at a finite number of

interest rates. Therefore, we use interval regressions to estimate subjects’ reservation interest 16 Although it is not the focus of their paper, Hilary and Hui (2009) find suggestive evidence that priming religion increases risk aversion in a sample of 120 undergraduates. It is difficult to directly compare their results with ours because they report only full-sample results and not the religious composition of their sample, they report only p-values (.20 and .08 for their two regressions) rather than effect sizes, they use a different priming instrument, and their subjects made only two risk choices that are both hypothetical.

Page 16: Religious identity and economic behavior 2010.04.13

16

rates. In the interest rate regressions that follow, if the coefficients imply that a certain set of

explanatory variable values is associated with a mean log continuously compounded interest rate

of μ̂ , then the median continuously compounded interest rate is ˆexp( )μ .

Table 3 presents the regression evidence on how priming religious identity affects

discount rates. The explanatory variables are a dummy for being in the religion-salient condition

and a dummy for the intertemporal choice being between payments deferred for one week versus

two weeks. Contrary to the hypothesis that religious identity promotes thrift and capital

accumulation (Weber, 1930; Guiso, Sapienza, and Zingales, 2003; Barro and McCleary, 2003

and 2006), in no case do we find that religious identity affects discount rates in a statistically

significant way. Moreover, the point estimates of the priming effect are positive for all religions.

E. Dictator game

Despite having incentives to keep their entire $1 endowment for themselves, dictators in

our dictator game usually gave away a positive amount of money to the subject with whom they

were paired, although the proportion given away was far less than half on average. This is a

typical result for dictator game experiments (e.g., Forsythe et al., 1994). Table 4 contains

coefficients from regressing the amount of money given away on a dummy for being in the

religion-salient condition. The constant terms indicate that when unprimed, the average amount

given away is between 14 and 22 cents, depending on the religious group.

The coefficients on the religion-salient dummy show that Protestant, Catholic, and Jewish

identities do not induce subjects to give away more money. The Protestant and Catholic priming

effect point estimates are in fact negative. Among atheists and agnostics, the priming effect is

negative and significant at the 10 percent level.

Thus, we fail to replicate the Shariff and Norenzyan (2007) finding that priming religion

increases generosity in a dictator game, even though we use the same priming instrument they

do. Shariff and Norenzyan do not report results separately by religion, but a rough meta-analysis

using full-sample results points to a precisely estimated zero effect overall. Assuming

independence between their Study 1 (β = 0.238, s.e. = 0.0645, N = 50), their Study 2 (β = 0.200,

s.e. = 0.0810, N = 50), and our experiment (β = -0.034, s.e. = 0.0214, N = 553), the combined

Page 17: Religious identity and economic behavior 2010.04.13

17

estimate of the priming effect on dictator game giving for a $1 endowment is $0.004 with a

standard error of $0.019.17

F. Labor market tasks

In the anagram solving task, our subjects on average supplied answers for around 30

anagrams during the five minutes allotted to them. But contrary to the hypothesis that religious

identity norms decrease the cost of work effort, the regression in Table 5 of anagrams attempted

on a dummy for being in the religion-salient condition shows that priming religious identity does

not significantly affect the number of anagrams attempted among any of our religious groups.

The point estimate of the priming effect on Protestants, who were the object of Weber’s (1930)

Protestant work ethic hypothesis, is in fact negative, although this is not significantly different

from zero. In untabulated analysis, we have used the number of anagrams correctly answered as

the dependent regression variable instead and find qualitatively similar results.

Labor supply effects in the bilateral gift-exchange game are summarized in Figure 1,

which plots the average amount Employees chose to expend on work-related costs for each wage

offer. Among all religious groups and experimental conditions, minimal labor is supplied at a $0

wage offer.18 The amount Employees are willing to expend on work-related costs rises nearly

linearly with wage offers between $0 and $3.50, indicating a strong reciprocity norm. In

measuring reciprocity, we will ignore labor supply at a $4 wage because at this wage, the

Manager’s earnings are always $0 regardless of how much labor the Employee supplies.

Therefore, any positive work-related expenditures by the Employee at a $4 wage represents pure

money burning as a gesture of goodwill and appreciation (or confusion about the payoff

formulas).19

17 To conduct this calculation, we divide Shariff and Norenzayan’s coefficients and standard errors by 10, since their subjects were told they were allocating a $10 endowment instead of a $1 endowment. In analyzing their Study 2, we ignore their second treatment group (which received a non-religious prime). Of course, differences in procedure or subject population may explain our non-replication. A potentially relevant difference is that their study involved deception; subjects’ choices were not actually implemented. Batson, Schoenrade, and Ventis (1993) review laboratory evidence that religion is related to self-reported helping behavior, but its relationship to actual behavior is much weaker. This pattern may help explain the divergent findings if incentive-compatibility was less credible to subjects in Shariff and Norenzayan’s experiment. 18 The absence of positive religious identity effects on purely altruistic labor provision in the gift-exchange game at a $0 wage is consistent with the lack of religious identity effects on pure altruism in the dictator game. 19 At a $4 wage, work expenditures are usually smaller than at a $3.50 wage. Primed Catholics appear to behave quite differently than other groups with respect to money burning. Unlike every other religion × experimental condition cell, primed Catholics slightly increase their work expenditures as the wage goes from $3.50 to $4. The

Page 18: Religious identity and economic behavior 2010.04.13

18

It is readily apparent from Figure 1 that the strength of reciprocity, as reflected in the

slope of work expenditures with respect to wages between $0 and $3.50, increases greatly among

primed Jews relative to unprimed Jews. Other religious groups’ reciprocity does not appear to be

nearly as affected by priming. We formally analyze the priming effect on Employees’ reciprocity

in Panel A of Table 6. The dependent regression variable is the subject-specific slope coefficient

from a regression of the subject’s work-related costs on wage offers from $0 to $3.50. Jewish

identity significantly increases this slope from 0.05 to 0.11, but there are no significant effects

for any other group.

Panel B of Table 6 examines the effect of priming religious identity on wages offered by

Managers. There is to our knowledge no clearly articulated hypothesis in the prior literature

about religion’s effect on managerial wage offers, so we report these results mainly for

completeness.20 The regression of managerial wage offers on a dummy for being in the religion-

salient condition shows no significant effects of religious identity.

G. Multiple hypothesis testing and Type I error

Although all of the main hypotheses we have tested correspond to existing hypotheses in

the literature, the number of tests we have run is large, and many of the priming effects are

insignificant. This raises the concern that our significant priming effects are chance artifacts

arising from the large number of tested hypotheses.

To address this possibility, we test whether the priming effects of each religion on public

good contributions, expectations of others’ public good contributions, risk aversion, discount

rates, dictator game generosity, anagrams attempted, and gift-exchange reciprocity are jointly

equal to zero.21 We reject this hypothesis at p = 0.006.22

difference between primed and unprimed Catholic work expenditures at a $4 wage is not statistically significant, however. 20 However, since one motivation for a Manager to pay a positive wage is that he or she expects the Employee to reciprocate the kindness, one might expect that hypotheses about religious identity effects on trust in the public goods context would also apply to Managerial wage offers. Our results suggest Catholic norms about trust are specific to public good contexts. 21 To run this test, we pool all of our observations into a single interval regression and control for outcome type dummies, outcome type dummies × religion dummies, and outcome type dummies × religion dummies × additional outcome-specific control variables. Interval regressions can accommodate dependent variables that are point observations as a special case. We allow the residual’s variance to vary by religion × outcome type and cluster standard errors by subject.

Page 19: Religious identity and economic behavior 2010.04.13

19

H. Treatment interactions with belief in divine punishment and religious service attendance

In this subsection, we examine whether the priming effects we have identified as

significant for Christians or Jews differ for subjects with a stronger belief in divine punishment

or more regular religious service attendance. Barro and McCleary (2003, 2006) find that the

positive association between religion and GDP growth appears to operate most strongly through

belief in the existence of hell, which could be a powerful motivator of behavior. In contrast, they

find a negative association between GDP growth and the frequency of religious service

attendance. Although Barro and McCleary interpret this negative relationship as arising from

religious service attendance being a proxy for real resources being diverted to religious activities

rather than economic production, it is possible that frequent attendees have different norms than

infrequent attendees, even holding beliefs about the afterlife fixed. Differences in priming effect

sizes across individuals could be due to differences in norms. But in Section I, we discussed

other reasons why priming effect sizes could vary, so we interpret these treatment interactions

with caution.

Our debriefing questionnaire asked a subset of our subjects to rate on a six-point Likert

scale their agreement with the statement, “God punishes people for their sins.” We normalize this

variable so that within each religious group, it has a zero mean and unit variance. We also asked

all subjects how often they attend religious services. Possible answers were “never,” “less than

once a month,” “once a month,” “a few times a month,” “once a week,” “a few times a week,”

“once a day,” and “more than once a day.” We create an indicator variable for whether the

subject’s attendance frequency is above the median for his or her religious group. Median

attendance frequency is once a month for Protestants, less than once a month for Catholics and

Jews, and never for atheists and agnostics.

Table 7 shows regressions where the explanatory variables include an interaction of the

religion-salient dummy with either the strength of belief in divine punishment or with the

indicator for greater-than-median frequency of religious service attendance.23 To keep the

22 We exclude the priming effect on Managers’ wage offers from this joint hypothesis test because there is no strong ex ante hypothesis from the literature about how religion would affect this dependent variable. If we include it, the p-value for the joint hypothesis test is 0.015. 23 Interpreting the interaction coefficient could be problematic if the prime affected self-reports of belief in divine punishment or frequency of religious service attendance. We find no systematic differences in either of these variables across treatment and control groups, with the exception that primed Jews report substantially higher belief

Page 20: Religious identity and economic behavior 2010.04.13

20

number of interaction regressions manageable, we limit attention to the dependent variables and

religious groups where we found statistically significant main effects of priming. For the sake of

brevity, we omit from the table the regressions with Catholic trust as the dependent variable,

since these results are similar to those from the regressions with Catholic public good

contributions as the dependent variable. We also omit regressions involving agnostics and

atheists because interactions with religious belief or religious service attendance are difficult to

interpret for this group.

There is a marginally significant interaction for Protestants’ contribution to the public

good. Protestants with the mean belief in divine punishment marginally increase their

contribution from 52 cents when not primed to 66 cents when primed. For Protestants with a

belief in divine punishment one standard deviation above the mean, the treatment effect appears

to be 26 cents instead of 14 cents. This effect, if real, is consistent with Barro and McCleary’s

(2003, 2006) hypothesis that the salutary effects of belief in divine punishment operate through

changes in norms.

We find no statistically significant interactions between attendance frequency and

religious identity effects among Protestants and Catholics. Among Jews, we find that the

increase in Employee reciprocity in response to the religious prime occurs primarily among those

who attend religious services relatively infrequently.24

IV. Conclusion

The debate about religion’s effect on economic outcomes has been hindered by the

difficulty in identifying exogenous variation in religion. In this paper, we created exogenous

variation by experimentally manipulating the salience of religious identity in laboratory subjects.

The long-standing psychological theory of self-categorization predicts that norms associated with

an identity have a temporarily greater behavioral influence when that identity is salient.

in divine punishment (p < .05). However, we find that the coefficients on belief in divine punishment and its interaction with the treatment dummy are essentially equal to zero in the gift-exchange reciprocity regression among Jews in Table 7. 24 Because we varied the order of experimental games and preference elicitations across sessions, another question we can examine is how the strength of the priming effect varied with the length of time elapsed since the priming manipulation. The effects on Jewish gift-exchange reciprocity and Catholic risk aversion may weaken over time, while the effects on Protestant and Catholic public goods contributions and Catholic trust appear to strengthen over time, but we put little weight on these possible trends because very few of the interactions between the priming effect and task order are statistically significant.

Page 21: Religious identity and economic behavior 2010.04.13

21

Therefore, we can identify the marginal directional effect of religious identity norms on

economic choices by seeing how those choices change when religious identity salience varies

exogenously.

We find that Protestantism increases contributions to public goods. Catholicism decreases

contributions to public goods, expectations of others’ contributions to public goods, and

decreases risk aversion. Judaism increases labor market reciprocity. However, we find no

evidence that religious identity affects discount rates or purely altruistic generosity.

Page 22: Religious identity and economic behavior 2010.04.13

22

References Akerlof, George A., and Rachel E. Kranton. 2000. “Economics and Identity.” Quarterly Journal

of Economics 115(3): 715-733.

Anderson, Lisa R., and Jennifer M. Mellor. 2009. “Religion and Cooperation in a Public Goods Experiment.” Economics Letters 105, 58-61.

Ashraf, Nava, Dean S. Karlan, and Wesley Yin. 2006. “Tying Odysseus to the Mast: Evidence from a Commitment Savings Project in the Philippines.” Quarterly Journal of Economics 121(2): 635-672.

Barro, Robert J., and Rachel M. McCleary. 2003. “Religion and Economic Growth.” American Sociological Review 68: 760-781.

Barro, Robert J., and Rachel M. McCleary. 2006. “Religion and Economy.” Journal of Economic Perspectives 20: 49-72.

Barsky, Robert B., F. Thomas Juster, Miles S. Kimball, and Matthew D. Shapiro. 1997. “Preference Parameters and Behavioral Heterogeneity: An Experimental Approach in the Health and Retirement Study.” Quarterly Journal of Economics 112(2): 537-579.

Batson, C. Daniel, Patricia A. Schoenrade, and W. Larry Ventis. 1993. Religion and the Individual: A Social-Psychological Perspective. New York: Oxford University Press.

Becker, Sascha O., and Ludger Woessmann. 2009. “Was Weber Wrong? A Human Capital Theory of Protestant Economic History.” Quarterly Journal of Economics 124: 531-596.

Benjamin, Daniel J., James J. Choi, and A. Joshua Strickland. Forthcoming. “Social Identity and Preferences.” American Economic Review.

Benz, Matthias, and Stephan Meier. 2008. “Do People Behave in Experiments as in the Field?—Evidence from Donations.” Experimental Economics 11: 268-281.

Bickel, Warren K., Amy L. Odum, and Gregory J. Madden. 1999. “Impulsivity and Cigarette Smoking: Delay Discounting in Current, Never, and Ex-Smokers.” Psychopharmacology 146(4): 447-454.

Carpenter, Jeffrey, and Erika Seki. Forthcoming. “Do Social Preferences Increase Productivity? Field Experimental Evidence from Fishermen in Toyoma Bay.” Economic Inquiry.

Charness, Gary, Guillame R. Frechette, and John H. Kagel, 2004. “How Robust is Laboratory Gift Exchange?” Experimental Economics 7: 189-205.

Dehejia, Rajeev, Thomas DeLeire, Erzo F. P. Luttmer, and Josh Mitchell. 2009. “The Role of Religious and Social Organizations in the Lives of Disadvantaged Youth.” In An Economic Perspective on the Problems of Disadvantaged Youth. Jonathan Gruber, ed. Chicago: University of Chicago Press.

Page 23: Religious identity and economic behavior 2010.04.13

23

Dohmen, Thomas, Armin Falk, David Huffman, Uwe Sunde, Jürgen Schupp, and Gert G. Wagner. 2005. “Individual Risk Attitudes: New Evidence from a Large, Representative, Experimentally-Validated Survey.” IZA Discussion Paper No. 1730.

Ellison, Christopher G. 1991. “Religious Involvement and Subjective Well-Being.” Journal of Health and Social Behavior 32: 80-99.

Fehr, Ernst, and Andreas Leibbrandt. 2008. “Cooperativeness and Impatience in the Tragedy of the Commons.” IZA Discussion Paper No. 3625.

Fehr, Ernst, Erich Kirchler, Andreas Weichbold, and Simon Gächter. 1998. “When Social Norms Overpower Competition: Gift Exchange in Experimental Labor Markets.” Journal of Labor Economics 16(2): 324-351.

Fischbacher, Urs. 2007. “z-tree: Zurich Toolbox for Ready-Made Economic Experiments.” Experimental Economics 10: 171-178.

Forehand, Mark R., Rohit Deshpandé, and Americus Reed II. 2002. “Identity Salience and the Influence of Differential Activiation of the Social Self-Schema on Advertising Response.” Journal of Applied Psychology 87(6): 1086-1099.

Forsythe, Robert, Joel L. Horowitz, N. E. Savin, and Martin Sefton, 1994. “Fairness in Simple Bargaining Experiments.” Games and Economic Behavior 6: 347-369.

Frederick, Shane, George Loewenstein, and Ted O’Donoghue. 2002. “Time Discounting and Time Preference: A Critical Review.” Journal of Economic Literature 40(2): 351-401.

Freeman, Richard B. 1986. “Who Escapes? The Relation of Churchgoing and Other Background Factors to the Socioeconomic Performance of Black Male Youths from Inner-City Tracts.” In The Black Youth Employment Crisis. Richard B. Freeman, ed. Chicago and London: University of Chicago Press, pp. 353-376.

Friedrichs, Robert W. 1960. “Alter versus Ego: An Exploratory Assessment of Altruism.” American Sociological Review 25(4): 496-508.

Froot, Kenneth A. 1989. “Consistent Covariance Matrix Estimation with Cross-Sectional Dependence and Heteroskedasticity in Financial Data.” Journal of Financial and Quantitative Analysis 24(3): 333-355.

Fuchs, Victor. 1982. “Time Preference and Health: An Exploratory Study.” In Victor R. Fuchs, editor, Economic Aspects of Health: 93-120. Chicago: University of Chicago Press.

Gächter, Simon, Benedikt Herrmann, and Christian Thöni. 2004. “Trust, Voluntary Cooperation, and Socio-Economic Background: Survey and Experimental Evidence.” Journal of Economic Behavior & Organization 55(4): 505-531.

Gruber, Jonathan H. 2005. “Religious Market Structure, Religious Participation, and Outcomes: Is Religion Good For You?” Advances in Economics Analysis and Policy 5(1), Article 5.

Page 24: Religious identity and economic behavior 2010.04.13

24

Guiso, Luigi, and Monica Paiella. 2008. “Risk Aversion, Wealth, and Background Risk.” Journal of the European Economic Association 6(6): 1109-1150.

Guiso, Luigi, Paola Sapienza, and Luigi Zingales. 2003. “People’s Opium? Religion and Economic Attitudes.” Journal of Monetary Economics 50: 225-282.

Hilary, Gilles, and Kai Wai Hui. 2009. “Does Religion Matter in Corporate Decision Making in America?” Journal of Financial Economics 93: 455-473.

Iannaccone, Laurence R. 1998. “Introduction to the Economics of Religion.” Journal of Economic Literature 36: 1465-1496.

James, William. 1890. The Principles of Psychology. New York: Henry Holt.

Kahneman, Daniel, Jack Knetsch, and Richard H. Thaler. 1986. “Fairness and the Assumptions of Economics.” Journal of Business 59: S285-300.

Khan, Uzma, and Ravi Dhar. 2006. “Licensing Effect in Consumer Choice.” Journal of Marketing Research 43: 259-266.

Kirby, Kris N., and Nancy M. Petry. 2004. “Heroin and Cocaine Abusers Have Higher Discount Rates for Delayed Rewards than Alcoholics or Non-Drug-Using Controls.” Addiction 99(4): 461-471.

Kirby, Kris N., Nancy M. Petry, and Warren K. Bickel. 1999. “Heroin Addicts Have Higher Discount Rates for Delayed Rewards than Non-Drug-Using Controls.” Journal of Experimental Psychology 128(1): 78-87.

Kumar, Alok, Jeremy Page, and Oliver Spalt. 2009. “Religious Beliefs, Gambling Attitudes, and Financial Market Outcomes.” University of Texas working paper.

La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert W. Vishny. 1997. “Trust in Large Organizations.” American Economic Review 87: 333-338.

Laury, Susan K., and Laura O. Taylor. 2008. “Altruism Spillovers: Are Behaviors in Context- Free Experiments Predictive of Altruism Toward a Naturally Occurring Public Good?” Journal of Economic Behavior and Organization 65: 9-29. LeBoeuf, Robyn A., Eldar Shafir, and Julia Belyavsky Bayuk. 2010. “The Conflicting Choices

of Alternating Selves.” Organizational Behavior and Human Decision Processes 111(1): 48-61.

Levin, Jeffrey S. 1994. “Religion and Health: Is There an Association, Is It Valid, and Is It Causal?” Social Science Medical Journal 24: 589-600.

Marwell, Gerald, and Ruth E. Ames. 1979. “Experiments on the Provision of Public Goods. I. Resources, Interest, Group Size, and the Free-Rider Problem.” American Journal of Sociology 84(6): 1335-1360.

Page 25: Religious identity and economic behavior 2010.04.13

25

Meier, Stephan, and Charles Sprenger. 2010. “Present-biased preferences and credit card borrowing.”American Economic Journal: Applied Economics 2(1): 193-210.

Petry, Nancy M., and Thomas Casarella. 1999. “Excessive Discounting of Delayed Rewards in Substance Abusers with Gambling Problems.” Drug and Alcohol Dependence 56(1-2): 25-32.

Putnam, Robert. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press.

Randolph-Seng, Brandon, and Michael E. Nielsen. 2007. “Honesty: One Effect of Primed Religious Representations.” International Journal for the Psychology of Religion 17: 303-315.

Reicher, Stephen, and Mark Levine. 1994. “Deindividuation, Power Relations Between Groups and the Expression of Social Identity: The Effects of Visibility to the Out-Group.” British Journal of Social Psychology 33(2): 145-163.

Reuben, Ernesto, Paola Sapienza, and Luigi Zingales. 2009. “Procrastination and Impatience.” Northwestern University working paper.

Rogers, William. 1993. “Regression Standard Errors in Clustered Samples.” Stata Technical Bulletin 13: 19-23.

Ruffle, Bradley J., and Richard Sosis. 2007. “Does it Pay to Pray? Costly Ritual and Cooperation.” The B.E. Journal of Economic Analysis & Policy 7(1) (Contributions), Article 18.

Sahm, Claudia R. 2007. “Stability of Risk Preference.” Federal Reserve Board Finance and Economics Discussion Series 2007-66.

Shariff, Azim F., and Ara Norenzayan. 2007. “God Is Watching You: Priming God Concepts Increases Prosocial Behavior in an Anonymous Economic Game.” Psychological Science 18: 803-809.

Shapiro, Jesse M. 2005. “Is There a Daily Discount Rate? Evidence from the Food Stamp Nutrition Cycle.” Journal of Public Economics 89(2-3): 303-325.

Stewart, Mark B. 1983. “On Least Squares Estimation when the Dependent Variable is Grouped.” Review of Economic Studies 50(4): 737-753.

Stulz, Rene M., and Rohan Williamson. 2003. “Culture, Openness, and Finance.” Journal of Financial Economics 70: 313-349.

Tan, Jonathan H.W., and Claudia Vogel. 2008. “Religion and Trust: An Experimental Study.” Journal of Economic Psychology 29, 832-848.

Page 26: Religious identity and economic behavior 2010.04.13

26

Toburen, Tina, and Brian P. Meier. Forthcoming. “Priming God-Related Concepts Increases Anxiety and Task Persistence.” Journal of Social and Clinical Psychology.

Turner, John C. 1985. “Social Categorization and the Self-Concept: A Social Cognitive Theory of Group Behavior.” In Advances in Group Processes, Vol. 2, ed. E. J. Lawler, 77-121. Greenwich, CT: JAI Press.

Weber, Max, 1930. The Protestant Ethic and the Spirit of Capitalism. London: Allen and Unwin.

Wheeler, S. Christian, and Richard E. Petty. 2001. “The Effects of Stereotype Activation on Behavior: A Review of Possible Mechanisms.” Psychological Bulletin 127(6), 797-826.

Page 27: Religious identity and economic behavior 2010.04.13

Table 1. Public Goods Game Results

Protestant Catholic Jewish Agnostic/AtheistPanel A. Amount contributed to public good

Religion salient 0.15** -0.18*** 0.01 0.12* (0.06) (0.07) (0.12) (0.06) Constant 0.53*** 0.70*** 0.56*** 0.49*** (0.05) (0.05) (0.08) (0.04) N 180 138 56 168

Panel B. Expectation of others’ contribution to public good Religion salient 0.03 -0.12** 0.08 0.09* (0.05) (0.05) (0.08) (0.05) Constant 0.59*** 0.73*** 0.61*** 0.52*** (0.03) (0.04) (0.05) (0.03) N 180 138 56 168

Panel C. Relationship between own contribution and expectation of others’ contribution Religion salient 0.12*** -0.07 -0.07 0.04 (0.04) (0.05) (0.09) (0.05) E(Others’ contribution) 0.90*** 0.94*** 1.01*** 0.95*** (0.05) (0.06) (0.11) (0.07) Constant -0.00 0.01 -0.06 -0.01 (0.04) (0.06) (0.06) (0.04) N 180 138 56 168 Note: This table shows regression results where the dependent variable is the amount contributed to the public good (Panels A and C) or the expectation of others’ average contribution to the public good (Panel B). Religion salient is a dummy for being in the religion-salient condition. E(Others’ contribution) is the subject’s expectation of other group members’ average contribution. Huber-White standard errors are in parentheses below the point estimates. * Significant at the 10% level. ** Significant at the 5% level. *** Significant at the 1% level.

Page 28: Religious identity and economic behavior 2010.04.13

Table 2. Risk Aversion Results

Protestant Catholic Jewish

Agnostic/ Atheist

Religion salient -0.03 -0.11* 0.02 -0.12** (0.06) (0.07) (0.06) (0.05) Larger Stakes 0.27*** 0.31*** 0.26*** 0.26*** (0.04) (0.05) (0.07) (0.04) Constant 0.23*** 0.21*** 0.09** 0.16*** (0.05) (0.05) (0.04) (0.03) σ̂ 0.39 0.36 0.26 0.32 (0.03) (0.03) (0.03) (0.02) N 242 154 80 196 Note: This table presents interval regressions where the latent dependent variable is the minimum risk premium required for a subject to accept a gamble. We pool each subject’s two risk choices together. Religion salient is a dummy for being in the religion-salient condition. Larger stakes is a dummy for if the sure payout in the risky choice was $100. The estimated conditional standard deviation of the latent dependent variable is denoted by ˆ.σ Huber-White standard errors, clustered by subject, are reported in parentheses below the point estimates. The final row shows the number of reservation risk premium intervals in the regressions. * Significant at the 10 percent level. ** Significant at the 5 percent level. *** Significant at the 1 percent level.

Table 3. Discount Rate Results

Protestant Catholic Jewish Agnostic/AtheistReligion salient 0.21 0.43 0.33 0.08 (0.36) (0.50) (0.57) (0.51) 1 week vs. 2 weeks 0.01 -0.17 -0.37 -0.30* (0.11) (0.17) (0.27) (0.17) Constant -2.96*** -3.53*** -3.62*** -4.40*** (0.26) (0.36) (0.44) (0.42) σ̂ 1.99 2.18 1.85 2.36 (0.16) (0.21) (0.20) (0.18) N 242 154 80 196 Note: This table presents interval regressions where the latent dependent variable is the log continuously compounded interest rate required to defer payment receipt. We pool each subject’s two intertemporal choices together. Religion salient is a dummy for being in the religion-salient condition. 1 week vs. 2 weeks is a dummy for the intertemporal choice being between payments deferred for one week versus two weeks. The estimated conditional standard deviation of the latent dependent variable is denoted by ˆ.σ Huber-White standard errors, clustered by subject, are reported in parentheses below the point estimates. The final row of each panel reports the number of discount rate intervals in the regressions. * Significant at the 10 percent level. ** Significant at the 5 percent level. *** Significant at the 1 percent level.

Page 29: Religious identity and economic behavior 2010.04.13

Table 4. Dictator Game Results

Protestant Catholic Jewish Agnostic/AtheistReligion salient -0.03 -0.04 0.01 -0.05* (0.04) (0.04) (0.07) (0.03) Constant 0.22*** 0.17*** 0.14*** 0.17*** (0.03) (0.04) (0.04) (0.02) N 186 139 62 176 Note: This table shows regression results where the dependent variable is the amount given away. Religion salient is a dummy for being in the religion-salient condition. Huber-White standard errors are in parentheses below the point estimates. * Significant at the 10 percent level. *** Significant at the 1 percent level.

Table 5. Number of Anagrams Attempted

Protestant Catholic Jewish Agnostic/AtheistReligion salient -1.48 1.76 4.92 2.12 (2.53) (3.98) (5.99) (2.63) Constant 29.80*** 31.09*** 33.65*** 32.02*** (1.89) (2.89) (4.48) (1.78) N 78 60 33 93 Note: This table shows regression results where the dependent variable is the number of anagrams attempted in the anagram-solving task. Religion salient is a dummy for being in the religion-salient condition. Huber-White standard errors are in parentheses below the point estimates. *** Significant at the 1% level.

Table 6. Bilateral Gift-Exchange Game Results

Protestant Catholic Jewish Agnostic/AtheistPanel A. Slope from regression of work cost on wages between $0 and $3.50

Religion salient 0.01 0.01 0.06** -0.02 (0.02) (0.02) (0.03) (0.02) Constant 0.08*** 0.08*** 0.05*** 0.09*** (0.01) (0.01) (0.02) (0.01) N 104 91 40 113

Panel B. Wage offered as manager Religion salient 0.02 0.01 0.42 -0.28 (0.19) (0.19) (0.29) (0.18) Constant 1.41*** 1.28*** 0.98*** 1.57*** (0.13) (0.14) (0.21) (0.12) N 104 91 40 113 Note: This table shows regression results where the dependent variable is the subject-specific slope coefficient from a regression of work cost expended as an Employee on managerial wage offered (Panel A) or the wage offered as a Manager (Panel B). Religion salient is a dummy for being in the religion-salient condition. Huber-White standard errors are in parentheses below the point estimates. ** Significant at the 5% level. *** Significant at the 1% level.

Page 30: Religious identity and economic behavior 2010.04.13

Table 7. Religion-Salience Treatment Interactions with Belief in Divine Punishment and Religious Service Attendance

Protestant Catholic Jewish Public good contribution Public good contribution Risk premium Gift-exchange reciprocity Religion salient 0.14* 0.14* -0.18** -0.20* -0.13 -0.15 0.06* 0.10*** (0.07) (0.08) (0.08) (0.10) (0.09) (0.09) (0.03) (0.03) Religion Salient × 0.12* -0.06 -0.08 0.01 Divine punishment (0.07) (0.08) (0.09) (0.03) Divine punishment -0.11** 0.04 0.06 -0.01 (0.05) (0.05) (0.06) (0.02) Religion Salient × 0.02 0.03 0.08 -0.16** (Attendance > median) (0.12) (0.14) (0.13) (0.07) Attendance > median -0.10 0.00 0.05 0.04 (0.09) (0.10) (0.10) (0.06) Larger stakes 0.38*** 0.30*** (0.08) (0.05) Constant 0.52*** 0.58*** 0.71*** 0.70*** 0.20*** 0.19*** 0.05*** 0.04** (0.05) (0.06) (0.06) (0.07) (0.06) (0.06) (0.02) (0.02) σ̂ 0.35 0.35 (0.03) (0.03) N 131 180 103 138 84 154 40 40Note: The column headings indicate the religious group and the dependent variable in the regression. The dependent variables are the amount contributed to the public good, the subject-specific slope coefficient from a regression of work cost expended as an Employee on managerial wage offered, and the minimum risk premium required to accept a gamble. The public good and reciprocity coefficients are from an OLS regression; the risk premium coefficients are from an interval regression where we pool each subject’s two risk choices together. Religion salient is a dummy for being in the religion-salient condition. Divine punishment is the self-reported belief in divine punishment, normalized to have zero mean and unit standard deviation within each religious group. Attendance > median is a dummy for whether the subject reports religious service attendance frequency that is above the median for his or her religious group. Larger Stakes is a dummy for the sure payout in the risky choice being $100. The estimated conditional standard deviation of the latent dependent reservation risk premium is denoted by ˆ.σ Huber-White standard errors are in parentheses below the point estimates; these are clustered by subject for the risk aversion regressions. The number of observations corresponds to the number of subjects for the public good and reciprocity regressions, and the number of reservation risk premium intervals observed for the risk aversion regressions.

Page 31: Religious identity and economic behavior 2010.04.13

Figure 1. Average Employee work costs chosen in response to managerial wage offers in gift exchange game.

$0.00

$0.10

$0.20

$0.30

$0.40

$0.0

0

$0.5

0

$1.0

0

$1.5

0

$2.0

0

$2.5

0

$3.0

0

$3.5

0

$4.0

0

Em

ploy

ee’s

wor

k co

st c

hose

n

Manager’s wage offer

Protestant

Control Religion salient

$0.00

$0.10

$0.20

$0.30

$0.0

0

$0.5

0

$1.0

0

$1.5

0

$2.0

0

$2.5

0

$3.0

0

$3.5

0

$4.0

0

Em

ploy

ee’s

wor

k co

st c

hose

n

Manager’s wage offer

Catholic

Control Religion salient

$0.00

$0.10

$0.20

$0.30

$0.40

$0.0

0

$0.5

0

$1.0

0

$1.5

0

$2.0

0

$2.5

0

$3.0

0

$3.5

0

$4.0

0

Em

ploy

ee’s

wor

k co

st c

hose

n

Manager’s wage offer

Jewish

Control Religion salient

$0.00

$0.10

$0.20

$0.30

$0.40 $0

.00

$0.5

0

$1.0

0

$1.5

0

$2.0

0

$2.5

0

$3.0

0

$3.5

0

$4.0

0

Em

ploy

ee’s

wor

k co

st c

hose

n

Manager’s wage offer

Atheist/agnostic

Control Religion salient